1. Field of the Invention
The present invention relates to an information processing apparatus for providing information.
2. Description of the Related Art
For provision of information by the Internet and broadcasts and provision of information stored in apparatuses in advance, the information is composed of one or more items which are written in a predetermined format. In general, an information providing apparatus provides a user with items necessary for provision of information by extracting the items from the information in accordance with the predetermined format. Concerning provision of information regarding a TV program, for example, electronic program information called EPG (electronic program guide) is written in a predetermined format. For each program, items such as a title, a subtitle, a performer, and a broadcast time are extracted from the EPG, and such items are provided to the user.
A technique is known in which information in the form of text to be provided is converted into synthesized speech information by speech synthesis, and the information is provided by means of synthesized speech. The information provision by means of synthesized speech eliminates the need for a step of displaying the information and draws a user's attention by means of sound, which are advantages of the information provision by means of synthesized speech when compared with the information provision by means of display.
In the information provision by means of synthesized speech described above, a large amount of information can take a considerable amount of time to be provided and provision of unnecessary information can unnecessarily distract a user's attention. To address these problems, attempts have been made to determine which information is to be provided by means of synthesized speech in accordance with attributes of an item, importance of contents of the information, or restrictions.
For example, Japanese Patent Laid-Open No. 05-165486 discloses a text-speech transforming device which has a table listing words forbidden from being broadcast to prevent phonetic output of words forbidden from being broadcast and which prevents read out of the words forbidden from being broadcast. Japanese Patent Laid-Open No. 09-288682 discloses an information selection speech output device in which information is composed of categorized items and only items having a keyword the same as that registered in advance are allowed to be read out.
When items of information that should be provided are apparent, text information of the items can be synthesized into speech and the speech is provided. However, when an item to be provided is not apparent, unnecessary information may be provided or necessary information may not be provided.
FIGS. 6A and 6B show examples of program information concerning programs. In FIGS. 6A and 6B, the program information is composed of eight items, that is, a start time, an ending time, a channel, a title, a subtitle, a performer, a category, and details. When a user requires information provision of a program title by means of synthesized speech, the information providing apparatus reads out an item corresponding to the program title. In FIG. 6A, the information providing apparatus can provide information regarding the title of a drama “Summer Vacation” as requested by a user. However, in FIG. 6B, since the title “Wednesday Cinema” does not specify the title of a movie, it is preferable to additionally provide a subtitle “King of Davide” by means of synthesized speech.
As described above, even if pieces of information are written in the same format, items which should be provided are not necessarily the same among the pieces of information. However, provision of all items increases the amount of information to be provided. In particular, since information provision by means of synthesized speech needs considerable time for reproducing synthesized speech information, reduction of the amount of information is required. In Japanese Patent Laid-Open No. 09-288682 mentioned above, items having a keyword are selected and subjected to speech synthesis for information provision. However, it is difficult to employ this method when there are a number of information names such as TV program names.